SMART BETA ROUNDTABLE: Convincing the sceptics

The appetite for smart beta strategies is growing, say our experts. but do investors understand how they work?

Smart_beta_roundtable_June_2017

David Wickham (senior investment specialist, Aberdeen Asset Management)
Chanchal Samadder (head of equity ETF strategy, Lyxor)
Hortense Bioy (director of passive funds research, Morningstar)
Manuela Sperandeo (iShares head of specialist sales, EMEA, BlackRock)

Funds Europe – Which type of investor provides the most demand for smart beta investing and how are other client channels developing?

Chanchal Samadder, Lyxor – If you go back to the 1970s and 80s, the first real users of these strategies were pension schemes institutions. The new vehicles like ETFs have brought these strategies more to the public attention, so now I think we’re seeing most demand for these types of products from asset managers and wealth managers.

David Wickham, Aberdeen Asset Management – Based on Spence Johnson’s 2016 European smart beta market intelligence report, using data from 2015, the European smart beta market is dominated by institutional investors (67%), such as pension funds, insurers, sovereign wealth funds, foundations, endowments and charities. Anecdotally, pension funds rather than other types of asset owners account for the majority of the total institutional market. Within the institutional channel, continental European institutional investors – especially in the Netherlands and Nordics – are the biggest adopters of smart beta. The wholesale/retail channel, comprised of private banks, discretionary wealth managers, advisers, platforms, etc make up the remainder (33%). Based on the US experience, where retail investors account for around half of smart beta assets (predominantly through ETFs), we do expect smart beta usage via ETFs to grow in Europe from its current low base.

Manuela Sperandeo, BlackRock – We tend to look at the experience of our US colleagues and see how the market has developed over there, where it is an increasingly wealth advisory phenomenon. In 2016 we started to see the first meaningful sign of these adoptions spreading out to advisers and retail and I suspect 2018 is going to be particularly relevant in that with MiFID II. We are seeing distributors getting ready for that and smart beta playing a role in becoming a key differentiator in terms of advisory offering, model portfolios.

Hortense Bioy, Morningstar – Dividend strategies have always been popular among all types of investors. Everyone wants income. Even retail investors were buying dividend strategies ten years ago, when smart beta wasn’t even called smart beta. In the last three years we’ve also seen intermediaries and private banks using minimum-volatility strategies. Dividend and minimum-volatility strategies are strategies that people easily understand, and that’s absolutely key. I think there’s still a fair amount of scepticism about smart beta, especially as the newer strategies are increasingly complex. Also, they don’t have a live track record.

Funds Europe – Can you name a couple of the strategies that may be receiving that scepticism?

Bioy – We’ve seen tremendous growth in multi-factor strategies over the past year or two, but people still need to understand what these are, because you can allocate across single factors, but you can also take a different approach and just add screens. For example, you add a minimum volatility screen or a quality screen to your dividend strategy. We consider these multi-factor strategies, but that is not the same thing as saying: ‘I am picking five factors and I am putting them together to create something that will smooth or mitigate the cyclicality of each factor.’ And when you do combinations like quality plus dividend, or minimum volatility plus dividend, it is easy to understand: you can break it down. But when you do value plus quality plus minimum volatility plus whatever, and you equal weight all these factors, or you opt for a dynamic allocation, then it becomes more like: ‘What are we doing there?’

Sperandeo – Hortense is right: the reason dividends and minimum volatility were the two exposures which had broader adoption from retail investors is because they are very straightforward in the outcome they deliver, which is income and lower risk. Yes, I think we need to do a far better job at articulating, every time we look at multi-factor combinations, what the outcomes that these combinations aim to deliver are going to be and what the instances in which they perform are going to be.

Wickham – Investors are increasingly sceptical of strategies like fundamental indexation which only provides an indirect exposure to factors like value and small size, as well as single factor strategies which can move sideways or underperform for significant periods.

As such, we believe multi-factor approaches are now preferred by investors, so we will soon be launching a proprietary and exclusive range of ‘Smarter Beta’ equity indices and funds that are truly multi-factor in approach.

Our approach aims to provide simultaneous positive exposures to all targeted risk premia, even when applied to a single factor product like value, for example – as we believe this to be superior to a single factor approach. This is a key point as many competing single factor designs have factor exposures that, more often than not, are underexposed to other desirable factors e.g. value often has a negative exposure to momentum, or even low beta, due to negative correlations. Given that all risk premia are engineered to provide risk-adjusted excess returns over the medium to long-term, we believe this to be a design flaw in competing designs and one that we have consciously corrected in our multi-factor approach.

Samadder – When you get to the topic of multi-factor it does become a lot more like active if you just think about the various potential outcomes, the ways you can combine the factors, top down, bottom up: how you build the individual factors. You get to the point where there is just the opportunity set and the spread of returns is going to be huge. Whereas with something like minimum volatility or single factor value, the definition is quite consistent, and whatever kind of vehicle you use, broadly your outcome is going to be roughly the same.

Wickham – Yes, there’s more scope for differentiation in multi-factor.

Funds Europe – It has been reported that there are over 300 factors. Is that right?

Samadder – I think if you go down to the real academic foundations, there are probably five to eight factors that most people agree on, and most of the other ‘factors’ are usually combinations.

Wickham –  I think the 300 or so supposed ‘factors’ are just signals, trading strategies or variations of the same core factors using different metrics. At Aberdeen, we have devised criteria called ‘RIPE’ – Robust, Intuitive, Persistent and Empirical – to determine which factors qualify as ‘risk premia’.

Robust means that factors must perform effectively in a constantly changing environment; Intuitive means that the factors should behave as expected and not be a ‘black box’; Persistent requires the factors to show persistence and generate excess returns over the medium to long term; and Empirical means that the factors should be based on, and supported by, empirical academic research. Within equities, this reduces the number of ‘risk premia’ to six – value, quality, momentum, small size, low volatility and options volatility – which we call ‘Ripe Factors’.

Sperandeo – There needs to be breadth in terms of these factors working across asset classes, but also for us it’s critical that these factors have a depth in terms of being implementable at scale and also very efficiently, because a  key consideration when it comes to factor investing is around maintaining this premium in terms of risk-adjusted excess return. The minute you start having poor implementation, for example by incurring excessive trading costs, this could really end up eroding all of your excess return.

Funds Europe – Is any method of implementing smart beta investing alongside traditional passive and active emerging to become the most prevalent?

Wickham – We view the asset management industry from an allocentric perspective which caters to a range of investor needs and requirements. This continuum ranges from high capacity, low-cost ‘beta’ strategies such as market capitalisation weighted indexation and enhanced indexation strategies through to lower capacity, higher-cost ‘alpha’ strategies.

Inbetween these is smart beta, a form of factor investing, which is increasingly seen by investors and their consultants as a third, non-market capitalisation approach that combines the best features of active and passive management.

Smart beta aims to achieve above-market returns or below market risk, or both, by gaining targeted exposure to ‘risk premia’ while retaining the numerous benefits of conventional indexing such as simplicity, objectivity, transparency, and relatively low costs.

Smart beta approaches are implemented through indices, which may entail: (1) the replication of third-party smart beta indices that are licensed in exchange for a licensing fee; (2) the creation of customised indices designed in-house by an investment management firm but independently calculated and constructed by a third-party index calculation agent, or (3) self-indexation ‘on the desk’ by an investment manager.

At Aberdeen, with over a decade of factor investing experience, we have opted for the second ‘customised index’ approach for external clients and will soon be offering a proprietary and exclusive range of eight ‘Smarter Beta’ multi-factor equity index families to showcase our capability and will launch investment products tracking the indices based on investor demand. 

Samadder – What we are seeing is clients use it on top of their portfolios as well to correct biases within the existing portfolio, so not either/or but looking at their entire portfolio, listing and looking at the factor exposure within that portfolio and saying: ‘Overall we are very underweight value with our active managers, but we like our active managers and we don’t want to take money away from them, but what we’ll do is overlay with the value ETF or other vehicle to just negate that bias from us.’

Funds Europe – Which is, or are, the most popular smart beta factors, and why? And how is the use of multi-factor investing developing?

Bioy – Dividend-screened/weighted strategies remain the most popular segment of the smart beta market in Europe, with a 43% market share. Again, the popularity of dividend strategies is explained by the appeal of income. A year ago, these were followed by low and minimum-volatility strategies. These were extremely popular last year to the point that people started worrying about their valuation levels and potential crowding. But low volatility is less popular now. Their market share has dropped from 19% last year to 12.7%. So, now the second most popular smart beta segment is multi-factor strategies, with a market share of 13.3%. In the second half of last year, value made a comeback after a decade of being the most unloved factor, so the market share of value has increased from 4.5% to over 10%.

Wickham – Fundamental indexation and single factors like value and low volatility have been popular in the past. However, we believe the industry is now at an inflection point and is starting to move towards multi-factor approaches. Indeed, the FTSE Russell 2017 smart beta survey shows that multi-factor is now the most popular smart beta equity strategy with 64% of asset owners evaluating multi-factor approaches (up from 37% in 2016 and 20% in 2015).

Sperandeo – Last year was the tipping point for the industry. Investors are opening up to other categories of smart beta strategies beyond traditional dividend and minimum volatility and while single factors are still the most favoured, multi-factor strategies are starting to gain traction, thanks to the longer live track record they now have.

Samadder – I think if you look at it over the long term, typically the most popular strategies have been value and dividend. I think what we’ve seen more recently is growing understanding and appreciation of other factors, and the question is how to rotate between them. What’s the most popular factor? It depends on where you are in the market cycle or what investors perceive is the market cycle.

Funds Europe – Are the smart beta inflows into mainly ETFs?

Samadder – Most of the strategies we run in smart beta are within an ETF wrapper, so for us it’s the first. I think globally across the industry it’s inside ETFs and outside ETFs.

Wickham – In the US, especially in the wholesale channel with registered investment advisors, for example, it is largely ETFs but in Europe it’s much more fragmented. Spence Johnson’s 2016 European smart beta market intelligence report shows the 2015 breakdown as 42% in segregated mandates, 41% in pooled funds and 17% in ETFs. So ETFs are the least popular at present but that’s off a low base and has a lot of room for future growth.

Funds Europe – Are there really more than 300 factors, according to a Research Affiliate paper, ‘A Smooth Path to Outperformance’, or is the industry overstretching the definition?

Bioy – With the computational power that you have at your fingertips, it’s kind of tempting to want to do better and find the next anomaly, the next pattern that make stock prices predictable. I think it’s human. I can see the temptation there. Now, investors need to be careful.

Wickham – It really depends on what you’re talking about. As mentioned before, if you’re talking about metrics used within factor design, there can be lots of different ways of doing the same thing. But if you’re talking about academically proven ‘risk premia’ then we believe it’s very difficult to justify more than six risk premia within equities.

Samadder – Going back to Hortense’s point, it’s very easy to just run a historic correlation and find correlation between two things and then attach some explainability to it and then call it a factor because it’s correlated to people without any causality, and I think that’s the risk.

Funds Europe – Can somebody give us a one-sentence definition on the difference between smart beta or strategic beta?

Samadder – I’d say factors are a sub-set of the broad smart beta, if you wanted one sentence.

Bioy – For us, smart beta or strategic beta is passive. What we mean by passive is that there must be an index, it’s just as simple as that. If it doesn’t track an index we don’t call it smart beta.

Wickham – There’s a lot of confusion in the industry around the definition of ‘smart beta’ and even the term itself. For instance, it is also referred to as scientific beta, advanced beta, alternative beta, alternative indexing, factor investing, amongst others. At Aberdeen, we define smart beta as non-market capitalisation, systematic (rules-based) investment strategies designed to deliver targeted exposure to risk premia that arise and persist in equity markets due to behavioural (non-risk based) and structural (risk-based) anomalies, with the aim of delivering superior risk-adjusted excess returns relative to equivalent market capitalisation weighted indices.

Funds Europe – Much smart beta has centred on equities, but a few firms, Deutsche Bank for example, have been talking about fixed income for a couple of years. Are other asset classes starting to open up?

Samadder – As soon as you go away from equities, liquidity becomes a very important factor. What we found is that when you apply those to credit or high yield, the transaction costs just eat away at any kind of enhancement the extra strategy itself delivers.

Sperandeo – I would argue in commodities, smart beta has been around for ages. In terms of swap strategies, they’ve always been about moving away from traditional production-based benchmarks, but they’ve not, I guess, been as adopted within ETF wrappers. In terms of the concept of moving away from traditional indexation and more into alternative ways of building benchmark indices,  commodities has been an asset class where this has happened for some time.

Wickham – We believe the next iteration is going to be ‘alternative risk premia’ strategies – those long-short approaches which capture both sides of the distribution and removes the beta component. A further iteration would be applying this across asset classes i.e. cross-asset alternative risk premia.

Samadder – Low cost’s something we’ve looked at, and I think again it comes back to the liquidity angle. If you want to do really proper long-short you want to be long factor and short anti-factor.

Academically that is the purest way of doing it, but trying to implement that again efficiently with the cost of borrowing and shorting stocks, it becomes very expensive. You don’t want to dilute it to be long factor and short market, because that’s not the real pure, pure outcome you wanted to live with.

Funds Europe – Is there a possibility active managers will one day stand accused of ‘factor hugging’, similar to the index-hugging controversy?

Samadder – Many clients develop their own kind of factor scoring tool in-house and are already analysing their active managers using the factor, which is probably the more relevant benchmark. Maybe not all active managers, but I think historically many have been guilty of tracking the wrong index, have been tracking global equities when they’re actually value managers and should be tracking global value index.

Wickham – Many active managers, either intentionally or unintentionally, are biased towards these factors anyway, such as value companies or small companies. With better technology and data analytics it is possible to discern those managers who are generating returns through idiosyncratic risk versus those who are simply benefiting from tilting their portfolios towards factors.

If active managers are largely generating returns from factors, then they really shouldn’t be charging active management fees as factors can be targeted and harvested more efficiently in a systematic manner at much lower cost. So I think the real issue is whether managers are being transparent.

Sperandeo – Now there is a very clear identified group of very sophisticated total large factor investors, who have completely moved from an asset class framework to a factor-based one. By speaking to institutional investors, we find that the journey to get there in many instances is to start with a dedicated factor allocation, which tends to become the benchmark against which measuring allocation to active managers. So I think it is increasingly becoming clear that your benchmark is no longer your market cap, it’s actually this family of smart beta indices and these are the indices active managers need to beat to demonstrate they are generating true alpha.

Bioy – There is an interesting question about those managers who say they are value managers, but then when you look at their portfolios, they regularly drift – and they’ve drifted a lot in recent years when value was not doing well. So, are you going to say these are bad managers?

If you know, as a manager, that you are going to be judged on how well you’ve done against a factor benchmark, then you can start with that universe. Then you will need to explain how you managed to deliver alpha.

Funds Europe – What are we going to be talking about next year? What will develop over the next 12 months?

Sperandeo – Next year is going to be all about timing. In 2016 we’ve seen the first big rotation in terms of flows and factors, away from minimum volatility into more cyclical factors such as value. The first five months of 2017 have been about value, so it’s going to be interesting to see what the next factor in favour is.

Wickham – Investors realise that it is possible to directly target desired factors. And given that factors are uncorrelated, or even negatively correlated, with each other since they tend to outperform at different stages of market and economic cycles, it is more optimal to blend factors to increase risk-adjusted returns. So I think we will still be talking about multi-factor in a year’s time.

Samadder – The next evolution would be if everything moves, once it’s here it would be probably cross asset multi-factor, equity risk premia, alternative risk premia, that kind of stuff as well.

Bioy – Maybe we’ll talk about performance as some of the strategies that came to market two or three years ago will have a track record.

Wickham – Investors now realise that it is possible to directly target desired factors. And given that factors are uncorrelated, or even negatively correlated, with each other since they tend to outperform at different stages of market and economic cycles, it is more optimal to blend factors in a multifactor approach in order to increase risk-adjusted returns.

So, I think we will still be talking about multifactor in a year’s time.

©2017 fund europe